Towards NMR analysis of the HIV-1 coreceptor CCR5 and its interaction with RANTES

CCR5 is a chemokine receptor together with CD4 used by HIV-1 as a primary gate of cell infection. For this reason CCR5 is of great interest for medicine as a target for the anti-HIV-1 therapies. Since the binding site of its endogenous ligand RANTES overlaps with the binding site of viral envelope glycoprotein gp120, a noninflamatory RANTES derivative 5P12-RANTES has been developed as an anti-HIV-1 infection microbicide. The primary aim of this thesis was to establish an NMR-amenable system to study CCR5 and to understand better the interaction with RANTES.
For this purpose CCR5 expressed in insect cells was characterized in detail with regards to its secondary structure, oligomeric state, particle size, stability, posttranslational modifications and functionality. In contrast to the previous results, carefully performed detergent screening revealed that FosCholine-12, a detergent which allows high yield purification, does not support CCR5 recognition by 2D7 and cannot be used for studying CCR5 interactions with ligands. Therefore for the functional studies the receptor was solubilized with a milder detergent mixture DDM/CHAPS/CHS, which was shown to support native CCR5 tertiary structure.
Using this setup it could be shown by SPR that 5P12-RANTES binds with higher affinity than another potent RANTES variant PSC-RANTES (Morin et al., manuscript in preparation). This explains why 5P12-RANTES, which unlike PSC-RANTES does not cause CCR5 internalization, is an equally effective anti-HIV-1 microbicide. On the other hand, the wild-type RANTES was shown to aggregate on the receptor micelle using a mechanism compatible with the linear oligomerization, a process that is proposed to serve local chemokine preconcentration.
To obtain an access to a cost-efficient source of isotope-labeled samples, an E. coli expression system was established for CCR5 (Wiktor et al., 2012, J Biomol NMR, in revision). The expression was facilitated by fusing the N-terminus of CCR5 to well expressing protein domains e.g. thioredoxin. The C-terminal CCR5 truncation and the mutation of cysteines increased the protein yield up to 10 mg/L and improved the sample stability. Due to the engineered thrombin proteolytic site the N-terminal fusion partner i.e. thioredoxin could be quantitatively cleaved and removed by size exclusion chromatography. The FC-12-purified receptor was abundant in alpha-helical secondary structure but could bind RANTES, MIP-1beta and conformation-dependent antibody 2D7 only when solubilized by a DDM/CHAPS/CHS mixture. Using 15N,13C,2H-labeled CCR5 2D and 3D NMR experiments were recorded but only about 80 backbone resonances could be resolved. The spectral quality was jeopardized by large overlap and line-broadening and needs further improvements to allow the assignment and the structural investigation.
To study 5P12-RANTES by NMR the backbone assignment was completed. The HSQC spectrum revealed that, unlike wild-type RANTES and other chemokines, 5P12-RANTES does not form dimers. The secondary chemical shift analysis suggest that the overall structure of 5P12-RANTES is similar to the wild-type RANTES, with the exception of the mutated N-terminus, which does not participate in the intermolecular beta-sheet and was shown to be highly flexible. Another important observation was that RANTES secondary structure is perturbed by Fos-Choline detergents, whereas maltosides shift the RANTES monomer:dimmer equilibrium towards its monomeric form.
The last part of the thesis present an independent study, where using ubiquitin as an example the mechanism of protein unfolding is studied (Vajpai et al., 2012, Proc Natl Acad Sci USA, in revision) manuscript submitted for publication). The secondary chemical shift analysis showed that the alcohol-denatured ubiquitin structure closely resembles the cold- and pressure denatured structure. This suggests that alcohol, low temperature and pressure unfold proteins by reducing the hydrophobic effect, the cost of exposing hydrophobic residues.
The data of this thesis will be presented in the following publications:
1. Wiktor, M., Morin, S., Sass, H-J., Kebbel, F., Grzesiek, S. (2012) Biophysical and structural investigation of bacterially expressed and engineered CCR5, a G protein-coupled receptor. J Biomol NMR (2012, in revision).
2. Vajpai, N., Nisius, L., Wiktor, M., Grzesiek, S. (2012) High pressure NMR reveals close similarity between cold and alcohol protein denaturation due to a reduction of the hydrophobic effect. Proc Natl Acad Sci USA (in revision).
3. Morin, S., Wiktor, M., Sass, H-J., Hartley, O., Grzesiek, S. (2012) Modulation of RANTES binding to CCR5 by modifications in the N-terminus and C-terminus (in preparation).